Measurement techniques for fresh concrete

As discussed earlier, a test characterizing the flow of concrete should be able to determine at least two parameters, such as yield stress and plastic viscosity. The design of a rheometer for concrete needs to take into account the dimensions of the coarse aggregate. The smallest gap in the instrument should be at least 3 times the largest diameter of the coarse aggregate to obtain a representative sample and to avoid the interlocking of the aggregates that will prevent flow. The difficulty in meeting this requirement led to the design of empirical tests that do not allow for the calculation of the yield stress and plastic viscosity in fundamental units. The design of such tests was to imitate the method of placement in the field. These tests very often measure only one value which is not necessarily related to the fundamental parameter defined by Bingham. It is only recently that some instruments were designed to obtain two values that are related to the fundamental parameters.

There are numerous standard and non-standard empirical tests to measure the flow of concrete. Because the results of such tests are not expressed in fundamental units, it is difficult to relate results from different tests. They can only be used for a direct comparison between concretes when using the same test.

As a full description of all the tests is beyond the scope of this review, we will limit ourselves to a list of the tests with some comments. There are two broad categories of tests: those that provide one parameter and those that provide two.

Table 1 gives a list of most common tests with some comments on the type of result that can be obtained. A discussion of the merits and results obtained can be found elsewhere [11 , 12]. To measure the viscosity, the yield stress needs to be exceeded. This can be achieved by various methods but the two most common are gravity or vibration. In the "gravity" method the stress applied is the weight of the materials, as opposed to an external applied stress. In the vibration method, the yield stress and flow behavior of the concrete are completely different from those observed without vibration. These tests are intended to simulate field performance in the laboratory.

Table 1: Tests that measure only one parameter, either yield stress or viscosity

A coaxial rheometer is composed of two concentric cylinders. The outer cylinder is usually stationary and the inner cylinder rotates at a controlled speed. The shear stresses generated by the fluid are measured on the inner cylinder. To be able to compute the shear stress and shear rates as well as calculate the yield stress and plastic viscosity according to the Bingham equation, the gap between the cylinders needs to be relatively small as compared to their diameters. It is generally accepted that the ratio of the radii of the two cylinders should be between 1 and 1.1. For concrete, the gap needs to be at least 3 to 5 times the size of the coarse aggregate to avoid interaction between the aggregates and the walls of the rheometer. Therefore, for an aggregate maximum size of 10 mm, the minimum radius are 0.5 m which will require the diameter of the outer cylinder to be between 0.53 and 0.55 m. These dimensions would have to be increased with the maximum size of the aggregate, rendering this type of instrument unsuitable for field use because it would not be easily transportable outside the laboratory. Such a rheometer was built by Coussot [25] and used for fresh concrete by Hu et al. [26] to validate the results obtained with the rheometer, BTRHEOM, developed at the Laboratoire Central des Ponts et Chaussées (LCPC).

To overcome the dimension limitations of the coaxial cylinder rheometer, while maintaining the possibility of estimating the two Bingham parameters, Tattersall [5] designed a rheometer that consisted of a shaft with blades that rotated in a bucket of concrete at a controlled speed. The torque generated by the concrete is measured on the shaft. This method does not allow for the calculation of viscosity and yield stress in fundamental units, but it enables the sudy of concrete flow under various shear rates. This rheometer, referred to as the "Two-point-test", was modified and computerized by Wallevik and Gjørv [27]. The commercially-available rheometer by Wallevik, BML [28], has another modification involving the shape of the blades. The blades are fins attached radially on the shaft. This rheometer can be used to estimate the Bingham parameters in fundamental units if no plug flow occurs [29]. Hu et al. [29] showed that plug flow occurs in concretes with a slump (measured according to the standards [13]) less than 200 mm. Beaupré [30] also developed a rheometer, referred to as IBB, with a different blade/shaft assembly, with the shape of the letter H. The rheometer IBB is also a modification of the original "Two-point test". Further descriptions of these tests are found elsewhere [12, 8].

Another geometry that is commonly used for rheological measurements is a parallel plate. Here, an upper plate rotates at a pre-selected speed and the torque generated by the shear resistance of the material is recorded on the same plate. The bottom plate is stationary. The shear rate in such an instrument is not constant and depends on the radial position, i.e., the shear rate is 0 at the center of the plate and maximum at the edge. In most cases the shear rate and the shear stress at the edge are the ones considered for the calculation of viscosity. This is not a serious problem if the fluid is Newtonian, because the viscosity does not depend on the shear rate, but it is for non-Newtonian fluids. For non-Newtonian fluids, an analytical calculation needs to be carried out. As before, the distance between the two plates has to reflect the size of the aggregates. This distance should be at least 3 to 5 times the diameter of the largest aggregate.

There is only one rheometer that uses the parallel plate geometry: the BTRHEOM [31] that was developed by de Larrard et al. at LCPC. It consists of a bucket that has a capacity of about 7 L, with a fixed wheel at the bottom and a wheel at the top rotating at any selected speed. The bottom wheel records the torque generated by the material reaction to shearing. The results of this test can be computed to obtain viscosity and yield stress in fundamental units.

Whereas, the concentric cylinder rheometers described above are too large for field use, the BTRHEOM is relatively small and can be carried by one person. Data acquisition is made with a portable computer.

Nevertheless, there is a need for a simple, inexpensive test to be used in the field for quality control of the concrete. A survey conducted by the National Ready-Mixed Concrete Association (NRMCA) in 1997 [32] showed that more than half of the participants indicated that although they considered the slump test adequate to describe workability, they felt that a better test was needed. They indicated that the slump test [13] did not give them a full description of the flow of concrete. For this reason, Ferraris and de Larrard developed at NIST a modified slump cone test [33, 34]. Figure 3 shows the schematic of this test. The modification consists of measuring not only the final slump height but also the time for the concrete to slump the first 100 mm, i.e., the speed of slumping. There are two methods to measure the speed:

• The original method consists of measuring the time for a plate to slide down with the concrete [34] a distance of 100 mm
• Researchers at Sherbrooke University [35] eliminated the plate and shorten the central rod so that its top was 100 mm below the full slump cone height. The test consists of measuring the time for the concrete to slump to the height where the rod first becomes visible.

The second method has the advantage that there is no risk of the plate getting stuck, but has the disadvantage that it may be difficult to see the appearance of the rod. Painting the end of the rod in a bright color does not solve the problem [35] because it is covered with cement paste.

From the final slump and slump time, the yield stress and plastic viscosity (in fundamental units) can be calculated using an empirical equation that was developed by comparing the modified test measurement with the values obtained with the BTRHEOM [9, 33 , 34].

This test is being evaluated in various laboratories in USA and France to determine the reproducibility of the results and the correlation between the slumping time and the final slump with the yield stress and plastic viscosity. When sufficient data has been collected, this test will be proposed to ASTM for consideration as a standard test.

In summary, while it can be seen that there are numerous tests to characterize the flow of concrete, few give results in fundamental units and therefore the rheological properties of concretes measured using different tests cannot be directly compared. Recently, new tests for characterizing concrete using a more fundamental approach have been developed. While, not all researchers agree on which test is the most suitable for the wide range of concretes in use today, tests that can give results in fundamental units and that can be used on a construction site should be favored, because comparison between test results can be achieved.